Marantz VP-12S3 Specifiche Pagina 5

  • Scaricare
  • Aggiungi ai miei manuali
  • Stampa
  • Pagina
    / 7
  • Indice
  • SEGNALIBRI
  • Valutato. / 5. Basato su recensioni clienti
Vedere la pagina 4
To Home Theater indicates a delay in the
blue channel. But the latter could also be a
result of shortening the leading edge transi-
tions. In any event, I couldn’t identify any
related image degradation or color fringing
artifacts on DVD movies.
You can spend your rainy afternoons
individually optimizing these adjustments for
your DVD collection, or you could try the
default settings that are loaded when you
select one of the predefined Picture modes.
I found all of the default Picture mode set-
tings to be much too aggressive, but you
might want to start with them and use the
Sharpness control to modify the overall pro-
cessing gain. For most of my DVD collec-
tion I ended up setting the Sharpness con-
trol to ‘1’ with very small individual enhance-
ment gain settings, or at zero to bypass the
enhancement. But there were a few DVDs
with rather soft transfers that I used some
additional enhancement to an advantage.
The Luminance and Chrominance
enhancement processing is not available for
analog RGB or DVI signals. For those inputs
a separate Sharpness Filter is added to the
Miscellaneous menu. It has 5 settings (0 to
4) where ‘0’ produces no edge enhancement.
The Delay control in the Chrominance
submenu was not useful. It is intended to
delay the color signals relative to the luma
signal, but it also separated the luma signal
into multiple color components creating an
effect that resembles misconvergence on a
CRT projector. Hence the Delay control should
always be left at its zero (default) position.
Standard-Definition
Scaling
Scaling performance varies depending on
the input signal type and format. The analog
480i YPbPr scaling is excellent. With the
Sharpness control turned off (zero) outlining
around horizontal and vertical edges is only
about 1 to 2 native (720p) pixels, and there
is no additional visible ringing. The 6.75
MHz burst (single-pixel black and white ver-
tical lines) from the AccuPel generator had
negligible bandwidth loss and little aliasing.
Overscan varied between 2 to 2.5 percent
top and bottom, and 3 to 3.5 percent on the
sides. The analog RGB input also accepts
480i signals, but the deinterlacing is poor
(not done by the Faroudja components) and
I wouldn’t recommend using it.
The appearance of 480p YPbPr signals
was not as good. Outlining around horizon-
tal edges was just 1 to 2 pixels, as it was
with 480i signals. But there were 4 to 5
(720p) pixels of outlining on vertical edges
and additional ringing when fed fast signal
transitions from the generator. I don’t
believe this was a scaling effect, but rather
an issue in the analog 480p YPbPr signal
path, because 480p RGB signals looked
very good. The fast signal transitions are
typical of a PC video card. When the AccuPel
generator’s slower edge transition option was
used, the outlining was reduced but still visible.
The top burst in the multiburst pattern
showed only a slight roll-off in frequency
response. Overscan was 3.5 percent on the
sides, and 2.5 percent top and bottom.
The DVI digital video input had no prob-
lem with 480p signals and provided excep-
tionally good scaling with the Sharpness
Filter set to zero. There were only 1 to 2
(720p) pixels of outlining on horizontal and
vertical edges, and all of the bursts in the
multiburst pattern looked excellent. In addi-
tion there was no overscan—not a pixel was
lost at the top, bottom, or sides of the image.
The S-video chroma bandwidth is excel-
lent. The 1.5 MHz chroma stripes on Video
Essentials’ Snell & Wilcox test pattern were
bright, although not equivalent to the fre-
quency response of the YPbPr signals. So
you should still use YPbPr signals from a
DVD player or a digital broadcast receiver
when they are available.
High-Definition Scaling
The projector produces a spatially “pixel
perfect” image when its DVI input is driven
by 720p video signals from an AccuPel
Calibration Generator. Each active (visible)
pixel from the source is mapped to a single
projector pixel. There is not a single pixel of
overscan or underscan. The single pixel
lines within a multiburst pattern are precise-
ly rendered as single pixel lines with full res-
olution. The Sharpness Filter is disabled
when the DVI (or analog RGB) input
receives 720p signals.
The DVI input also worked well for 1080i
digital signals. There were only 1 to 2 pixels
of faint outlining on vertical and horizontal
lines in the Sharpness pattern, with 1 to 2
pixels missing at the edges. However, the
combined effect of the 1080i deinterlacing
(vertical-interpolation) and scaling could be
seen on the overscan patterns, where some
of the horizontal lines were 3 to 4 pixels
thick. The Sharpness Filter had little effect
on the 1080i edges, but it affected the
mulitburst response. The top burst (1920
pixels per picture width) was necessarily
aliased, but the filter rolled off the aliasing
when set below 4. The default (2) setting
was best, below that the frequency
response of the previous burst also began
to roll off.
When the analog YPbPr input is used for
720p signals, the projector scales the
image to create about 1 to 1.5 percent over-
scan on each edge of the screen. The signals
are apparently filtered and produce no visi-
ble lines in the top two bursts of the
AccuPel 720p multiburst pattern, which
means that the horizontal resolution is
below 640 pixels per picture width. There is
also significant outlining (about 6 pixels) on
vertical edges and additional ringing.
Slower edges only slightly reduce the outlin-
ing artifacts. The analog YPbPr input per-
forms similarly with 1080i signals. There are
4 to 5 pixels of outlining, and the overscan
is about 1 percent on each edge. Again
there are no visible lines in the top two
bursts of the multiburst pattern, which
means that the horizontal resolution is
below 960 pixels per picture width.
Because of the poor frequency response, I
would avoid using the YPbPr inputs for
high-definition signals and use the DVI or
RGB input instead.
The analog RGB input does a much bet-
ter job with 720p and 1080i signals. When
the slower AccuPel edge mode is used,
which is equivalent to a high-definition set-
top box, the edge outlining is acceptably
faint. But when the faster edge mode is
used, which is similar to most PC video
cards, the edge outlining was about 8 pix-
els to the right of vertical edges on 720p
signals (5 to 7 pixels on 1080i) and there is
additional ringing. So I would avoid the ana-
log output of unfiltered PC video cards. The
720p single pixel lines in the top burst of
the AccuPel multiburst pattern showed
good resolution and depth of modulation.
The single pixel lines in the 1080i multiburst
were aliased. The Sharpness filter is
enabled for 1080i and the best setting was
the default (2) setting, as it was for 1080i
DVI signals. The Auto Adjust feature worked
well, and produced an image with no over-
scan or missing pixels for 720p. With 1080i
analog signals there was 2 percent top and
bottom overscan, 1 percent right edge, and
none on the left edge.
Deinterlacing
The projector uses Faroudja deinterlac-
ing technology, which includes film-mode
(inverse-telecine) deinterlacing for stan-
dard-definition movies, and DCDi
(Directional Correlational Deinterlacing) pro-
cessing for standard-definition original inter-
laced video.
Inverse-telecine deinterlacing provides
artifact-free deinterlacing of film-source
video as long as the video processor can
lock onto the 2-3 field pulldown cadence
that results from transferring 24 frame-per-
second film to 60 field-per-second inter-
laced video. It also completely eliminates
vertical interpolation in the deinterlacing
process, thereby maintaining the full resolu-
Equipment
Review
Widescreen Review • Issue 84 • May 2004
44
Page 5/7
#84 Master Pages 26-49 3/22/04 4:32 PM Page 44
Vedere la pagina 4
1 2 3 4 5 6 7

Commenti su questo manuale

Nessun commento